Train a Hotdog Image Recognition Model with Python
Learn to train a hotdog image recognition model using Python. Understand the process of image classification and model training with real-world datasets for building accurate image classifiers.
At a Glance
We all know that machines can do a lot these days, including recognizing whether or not an image has a certain object in it. But did you know that you, too, can train a model to do just that? In this guided project, you’ll learn how to train a model in Python with PyTorch, a machine learning library, to detect if a picture has a hotdog in it. This process can be repeated with any object, whether it is a bird, a plane or even Superman!
Have you ever wanted to create an image detector that would tell you whether or not a picture is of a certain object? Look no further because we have you covered!
In this project, we will be utilising simple learning rules to “teach” a neural network how to recognize hotdogs an image and solve the big question: hotdog or not hotdog?
Let’s find out!
A Look At the Project Ahead
Once you have completed, this project, you’ll be able to:
- Create Python functions with libraries to preview, load and train your model to recognize hotdogs with PyTorch
- Transform a dataset of images to prepare it for training
- Train a state of the art image classifier by using transfer learning: training a model to solve one problem and applying that training to a related problem
What You’ll Need
Before starting this lab, it’ll be helpful to be familiar with the following:
- Basic Python knowledge
- Basic knowledge about image classification and neural networks, provided in optional readings
If you’re interested in learning more about image classification and neural networks you can take a look at the following courses: Computer Vision and Image Processing Fundamentals course on Coursera or edX.
Instructor
Kathy An, IBM
Weiqing Wang, IBM
Weiqing Wang, IBM
Other Contributors
Joseph Santarcangelo, IBM
There are no reviews yet.